Researchers have recently published a study revealing a new, less invasive method for evaluating intracranial pressure (ICP) in patients.
The research, conducted by experts at Johns Hopkins Medicine, explores a potential alternative approach for measuring ICP.
The study, released on July 12 in the journal Computers in Biology and Medicine, highlights the significance of ICP, a crucial physiological parameter that can rise to abnormal levels following acute brain injury, stroke, or cerebrospinal fluid blockage. Elevated ICP can lead to symptoms such as headaches, vision problems, nausea, behavioral changes, and reduced consciousness levels. Monitoring ICP is essential for at-risk patients, but the current method involves highly invasive procedures like inserting an external ventricular drain (EVD) or an intraparenchymal brain monitor (IPM) directly into brain tissue through the skull.
“ICP is a vital sign in serious neurological conditions, but existing monitoring techniques are invasive, risky, and resource-intensive. We explored a new approach using Artificial Intelligence as a potential noninvasive method for assessing ICP,” explained Dr. Robert Stevens, director of anesthesiology and critical care medicine precision medicine.
Procedures like EVD pose risks of complications like catheter misplacement, infections, and bleeding, as demonstrated in recent research. These methods also require specialized equipment and expertise, creating a need for a safer alternative for ICP monitoring.
The research team at Johns Hopkins University School of Medicine, led by Dr. Stevens, hypothesized that severe brain injuries and increased ICP may impact cardiocirculatory function, leading to changes in systemic physiological patterns. They investigated the correlation between ICP patterns and three commonly monitored physiological waveforms in intensive care units: invasive arterial blood pressure (ABP), photoplethysmography (PPG), and electrocardiography (ECG).
Through advanced deep learning algorithms trained on ABP, PPG, and ECG data, the team achieved a level of accuracy in predicting ICP that surpassed existing methods. These findings suggest a promising noninvasive approach to ICP monitoring in patients.
Dr. Stevens emphasized the potential of AI-driven solutions to broaden access to ICP monitoring across diverse healthcare settings, pending validation of such physiology-based technologies.
Co-authored by several members, including students and faculty from Johns Hopkins University and the University of California San Francisco, the study marks a significant advancement in noninvasive ICP assessment.
